A Plug & Play AI Appliance Ready to Go for your Gen AI applications
Dell and Qubrid AI have partnered to bring you this unique, simple and powerful AI appliance! QAI-LLM-4DL40S is a plug and play system to help you get started and scale your journey with AI. Powered by powerful NVIDIA L40S GPUs, Dell hardware and systems management along with Qubrid AI Controller software – all you need to do it power and boot into the system and accelerate your AI projects without going through manual setup, management and operations of your AI infrastructure and tools. The system is designed to be simple and comes with key features as below:
Centralized AI Management & Enterprise Ready
Manage your entire data center GPU infrastructure and AI tools from a single platform, simplifying operations and reducing complexity.
- Start with a single appliance or deploy data center wide racked appliances
- Manage homogeneous or heterogenous server appliances featuring top NVIDIA GPUs – H200,, H100 and L40S etc
- Create enterprise account and assign nodes, users etc as per departmental or researcher needs
- Update OS and required NVIDIA GPU drivers seamlessly to the latest versions
- GUI based management for your favorite NVIDIA-SMI Command actions
- Single dashboard to view all your AI infrastructure and tools
Open Source AI Models out of the Box
Manage your interactions with popular AI models with intuitive user interfaces.
- Library of published Open-Source AI Models for tuning and inferencing available on-demand
- Fine-tune AI models on your local GPU server or scale across thousands of GPUs
NVIDIA NIM Microservices Integrated
Part of NVIDIA AI Enterprise, NVIDIA NIM is a set of easy-to-use inference microservices for accelerating the deployment of foundation models on any cloud or data center and helping to keep your data secure.
- Fully Integrated NVIDIA Enterprise catalog (NIM, CUDA, Nemo, etc.)
- Note – NVIDIA NIM requires separate Enterprise AI license – please contact us for more info.
Easily Deploy Hugging Face AI Models On Your GPU Appliance
The Qubrid AI Controller software allows you to easily deploy AI models of your choice from the Hugging Face repository. Simply enter the AI model ID, select number of GPUs and deploy the model. You can then do inferencing on these models on any GPU node in your infrastructure. Qubrid AI offers you the choice of curated open-source AI models, NVIDIA optimized NIM catalog or selection from thousands of models on Hugging Face – all deployable and manageable from the same software.
No-Code Fine-Tuning and RAG
Fine-tuning an AI model does not have to be hard. With our AI appliances, you don’t have to be a programmer or data scientist to fine-tune a model. same for RAG – just upload your departmental data and hit a button to take advantage of close to real-time RAG capabilities.
- Simple no-code fine tuning and RAG but with ability for advanced coding in Jupyter notebook
- Name your fine-tune model and save as templates
- Input multiple data types such as pdf files, images etc with a push of a button
Deep Learning & Machine Learning Packages Included
No more headaches of managing AI packages. Even with factory loaded packages, it is hard to manage and update these packages manually. The AI Controller automates that for you.
- One touch deployment of complete AI/ML Deep Learning packages (PyTorch, TensorFlow, Keras, etc.)
- Automated install and update for your discovery and research needs
- Continuous addition of new open-source tools
Hardware Specifications
This appliance is built using NVIDIA’s H100 GPU. Below are complete specifications:
- Dell PowerEdge R760XA
- 2U Rackmount form-factor
- 4 x NVIDIA L40S 48GB GPU
- 2 x 32-core or higher CPU
- 1TB of ECC Registered Memory
- 4 x 3.8TB NVMe 2.5” SSD
- 2 x 2-port 100Gb Network Ports (or 2 x 1-port 400Gb Network Ports)
- Dell Systems Management Software
World Class Hardware, Software and AI Support
This appliance comes default with our comprehensive 3-year next-day onsite hardware assistance, ensuring your systems remain operational and efficient. Additionally, our dedicated AI support covers everything from AI model consulting to CUDA help and fine-tuning assistance, providing you with the expertise you need to maximize your AI technology investments. With our commitment to customer satisfaction, you can rest assured that help is just a call or email away, every step of the way.
